Our work confirms and extends evidence that the cognitive performance of attention deficit disorder (ADD) children depends on the salience of the task; this dependence is predicted by a hypothesis that likens the ADD state to that of normals in "vigilance-decrement" after extended performance. We propose to assess ADD performance in terms of level and variability over time of (a) paired associate learning, (b) motor tracking, concurrently monitoring (c) gross motor activity. We shall determine whether the attention deficit chiefly involves (1) steeper performance decrement and greater increase in variability over time (the sustained attention model) or (2) impaired and excessively variable performance throughout (the vigilance decrement model). We shall also determine whether (3) stimulant-induced improvement shifts performance toward the normal pattern and (4) whether the performance improves monotonically across the full range of methylphenodate dose (.00 to .75 mg/kg), or peaks at an intermediate dose level, becoming impaired and more variable at higher doses. Applying the vigilance decrement model, we shall determine whether: (5) increasing tsk salience alters performance in the same manner as increasing dose; (6) salience manipulations, in combination, are additive; (7) salience and stimulant manipulations, in combination, are additive; and (8) positive salience response functions (with better performance at high levels) predict favorable stimulant response, and negative salience response functions predict adverse response. The above is a step toward a systematic basis for predicting an ADD child's favorable versus adverse response to stimulant medication, as it depends upon the test situation. In this manner, currently unexplained variability of ADD performance could be accounted for and potentially minimized. The outcome may suggest that stimulant effects should be monitored on measures spanning a range of stimulation levels, because an optimal dose for one measure may be excessive for another. Stimulant dosage could then be adjusted so as to maintain the best attainable balance between its divergent effects on different aspects of the child's life.